You are building Neuron23 in a challenging market. How does it affect how you do it?

When money and resources are scarce, you have to be especially careful where and how much you invest. You have to be prepared to trim, and to accelerate. You have to be critical and careful. You also have to be bold. Neuron23 wants to get to a place quickly with these therapies. So we took a stage-gated approach to deploying the capital we raised in 2021. We worked to be good stewards of that money and to stretch it out. We’d initially envisioned more programs. But we had to at least get our lead Parkinson’s program to patients. So we narrowed our scope. That’s helped us ride out some storms. But now we’re accelerating, because there’s a lot of promise in what we’re doing.

You’re on your fourth company. What have you learned along the way?

Companies rarely follow a straight line. Maybe you start off as an immunology company and end up as something different altogether. As a leader, you have to be open to that and surround yourself with talented people who are, as well. Having people around who are versatile, have complementary skill sets, and are willing to flex across different roles and disciplines is key. Staying close to the science has also been important to my success. I can run companies and bring medicines to patients. I’m not an MD, but I am a scholar of the science so that I can understand what we work on. As you go down a path in the development of new therapies, there are difficult calls to make sometimes: when to pivot, when to move, when to accelerate because things are promising. If you’re not close to it, you can make the wrong judgment.

Could your work have happened without AI and computational biology?

Could have? Possibly. Would have? Likely not. Could have, if we’d studied thousands of Parkinson’s patients over many years. And if we gave the new drug we’ve developed to thousands of people, which is atypical in a clinical trial and also expensive. We might’ve been able to reverse-engineer some meaningful results. Wouldn’t have, because that’s not how drug development works. The current process is long, slow, and expensive. What got us here is a significant volume of population-based data. We built a model of Parkinson’s disease with publicly available data shared with us by the Michael J. Fox Foundation. We put a huge amount of data into our ML model, which then got to work analyzing it much faster than any human could.

What did that large volume of data enable?

The volume allowed us to look — in a completely unbiased way — at the disease, and specifically one particular mutation. It’s known to cause 2–3% of Parkinson’s disease. Through the data analysis performed by ML , we found a genetic signature in another 30% of patients. The drug is currently in safety studies with healthy volunteers. Next year, we’ll start testing both groups of patients, in clinics. I can’t overstate the importance of the data we’ve been able to access. To build on the thesis that not every patient has the same exact mechanism of disease and same biology, you have to understand large numbers of patients and look at their disease in an unbiased way.

You’ve hinted at bias a couple of times. Can you say more about that?

Thinking we have it all figured out has led us astray in science, and in other fields too. There’s a lot of bias in the scientific community. It has been estimated that as many as half of scientific publications aren’t easily reproduced or replicable by an independent lab. You may get excited by a finding or trend. But you have to go back with fresh eyes and ask, “What are we seeing?” And be open to what the data tell you. A sense of wonder and curiosity will allow that. But you also need quality data. It’s the “garbage in, garbage out” concept. You need highly curated data as a good basis for your decisions.

Is the application of AI and ML in biotech fundamentally different from its application in other sectors?

In our case, we have an idea we want to test. We let ML and computational biology (with some programming from our side) look at data. But there’s a point where we lock in the information. We can’t keep going back to the FDA to say, “I’m using a different test this week. And next week, I’ll add 100 more data points and run another test.” In tech, finance, and other industries, you may have the same concept of wanting to test ideas. But there’s more ease in the continual evolution of models and real-time updating of data.

What have you learned about putting data at the center of your development process? What should other entrepreneurs take away from it?

Using big data can be daunting. We tackle this in two ways. We focus our questions. And we use high-quality data. We’re living proof that a small, focused team with finite resources can get to a powerful result. Now, accessing data isn’t always straightforward. We’re fortunate to have publicly available data sets that are rich and highly curated. We’ve also purchased and partnered to get data. I’ve also learned there may be cases where ML isn’t particularly helpful. You have to be open to that as well. It’s a great tool, but sometimes you don’t need it. And you shouldn’t force-fit it into a program.

What’s it like to be a woman leader in this field?

Early on, there are probably equal numbers of women and men in science. As you climb the ranks, that definitely changes. Far fewer of us are in higher-level positions, though it’s improving. So you don’t have as many peers or mentors. As a leader, there are benefits. There are people who want to work with me because I’m a woman. They’ve worked for men their whole lives. It’s neat to be singled out and have someone say, “I want a different experience.” Everyone brings something unique to leadership. I’m a very high-touch CEO. I can be very warm. And I can be very firm. Folks are drawn to me because I provide a unique mindset and perspective. Not better, not worse, but different.

English

Q&A

Nancy Stagliano: Bringing AI to life

Neuron23’s CEO talks about company building and using machine learning to develop individualized therapies

Companies rarely follow a straight line. As a leader, you have to be open to that, and surround yourself with talented people who are, as well.Nancy Stagliano, CEO, neuron23

You talked about the promise of what Neuron23 is doing. Can you explain? And how will it affect people’s diseases, their lives, and the industry at large?

We are driven by the idea of finding the right drug for the right patient at the right time in their disease. A better understanding of that has many benefits. It can change the course of disease in measurable ways. People will be able to say, “I can walk for a couple more years,” or “I can keep my cognition,” or “My ability to think will last longer.” That’s all possible if doctors intervene early with the right drug and slow down the disease process.

Then there are the dramatic costs of supporting the aging global population. These could be cut down if we got the right drug to the right patient, their disease did not progress, and they didn’t get debilitated from the wrong therapy.

We’re one of the first, if not the first, to bring this approach to neurodegeneration. If we can pave the way a bit, others will follow. People will recognize the value of precision medicine and a path to use it. That will make a big difference in our field in the way clinical trials are done and the way drugs are approved by the FDA.

I think we’ll see results of our work within the next decade. In our field, that’s fast.

Interactive Data

CFOs are cautiously optimistic about sidestepping a recession

Explore the full results of our Q2 2023 CFO survey. Compare and benchmark your company.

Feature

SambaNova: Dancing with AI

How SambaNova is making the most of an opportunity that’s materializing fast and at unprecedented scale